Handwritten Math Expression Recognition

Classification:

Python program to classify individual handwritten mathematical symbols from the CROHME isolated symbol competition data set. For classification, we used

1. a Support Vector Machine, and
2. a decision tree-based model(e.g., random forest, gradient boosted trees) provided by the python scitkit-learn library: https://scikit-learn.org/stable

Segmentation

Program that reads stroke data from a CROHME.inkml file, segments strokes into symbols, classifies the symbols, and then produces a label graph (.lg)file inObject-Relationship Formatas output.

Parsing

A system that takes CROHME stroke data, and produces a Symbol Layout Tree (SLT) as output. Such a representation can be translated directly to MathML, LATEX, andother representations for the appearance of math expressions.